dc.description.abstract
While Computer Tomography has become a standard tool in clinical biomechanics and biomaterials studies, much of the physical and chemical information contained in such scans remain unused. Instead, the grey values of the Computed Tomographs are, as a rule, directly related, by some empirical functions, to mass density or to mechanical properties of interest. However, a suitable combination of X-ray physics fundamentals, multiscale continuum micromechanics, and Finite Element analyses, as reviewed shortly in Chapter 1, indeed bears the potential for transforming the aforementioned, somehow hidden information, into tissue composition and microstructure, its heterogeneous distribution across investigated organs or implants, and the corresponding mechanical properties resulting from microscopic interaction of material constituents across several length scales. Chapters 2 to 5 contain corresponding examples for increasingly mature CT-to-mechanics conversion schemes, dealing with bone biomaterials and bony organs at different scales. Chapter 6 provides an outlook for extending such studies even to evolving biological systems, by combining trabecular bone micromechanics with advanced bone remodeling algorithms. In Chapter 2, the voxel-specific volume fractions of mineral, collagen, and water are derived from the measured X-ray attenuation information quantified in terms of grey values , by accounting for tissue-independent bilinear relations between mineral and collagen content in extracellular bone tissue (J. Theor. Biol. 287: 115, 2011). The aforementioned volume fractions enter a micromechanics representation of bone tissue, so as to deliver elastic properties in terms of voxel-specific stiffness tensors. The insertion of these properties into a FE simulation reveals that the choice of appropriate material properties influences the strain energy density in the extracellular matrix (governing the stiffness of the organ), and also affects the discretization level needed for obtaining converged numerical results. In Chapter 3, driving the field of Computed Tomography towards more quantitative, rather than qualitative, approaches, a new evaluation method is presented, which uses the unique linear relationship between grey values and X-ray attenuation coefficients, together with the energy-dependence of the latter, in order to identify the average X-ray energy employed in the scanner, and the nanoporosity of a tricalcium phosphate scaffold. This approach is extended in Chapter 4 by re-constructing the linear relation between the clinically accessible grey values making up a Computed Tomograph and the X-ray attenuation coefficients quantifying the intensity losses from which the image is actually reconstructed. Therefore, X-ray attenuation averaging at different length scales and over different tissues is combined with recently identified -universal- composition characteristics of the latter. This gives access not only to the normally non-disclosed X-ray energy employed in the CT-device, but particularly to in vivo, patient- and location-specific bone composition variables, such as voxel-specific mass density, as well as mineral and collagen contents. This is shown by example of a third lumbar vertebra. The corresponding vascular porosity values enter a continuum micromechanics model for bone (Ultrasonics 54:1251, 2014), which thereupon delivers voxel-specific elastic properties. The latter are mapped onto a 3D Finite Element mesh developed from the same patient data. The stress states resulting from corresponding Finite Element analyses are inputs for a six-scale strength upscaling model for bone, so as to compute element-specific proportionality factors to material yield or material failure. The implementation of patient-specific material properties highlights that simulations with averaged properties underestimate the fracture risk in bone, while the new approach reliably considers the effect of the material heterogeneities arising from bone remodeling triggered by everyday spinal loading; and is also relevant for even more heterogeneous, pathological cases. The last work, presented in Chapter 5, is using a multiscale analytical approach, which combines bone structural information at multiple scales to the remodeling cellular activities, and more precisely the mechanical stimulus sense by the osteocytes, in order to form an efficient, accurate, and beneficial framework for the prognosis of changes in bone properties due to aging or pathologies. This latter approach, once combined with the CT-based technique covered in Chapters 2 to 4, holds the promise to establish new forms of simulation-supported therapeutic activities in orthopaedy.
en